In today’s global marketplace, just one hour of unplanned downtime could cost your company hundreds of thousands of dollars. That’s why chances are good that you are being asked to ensure the availability of your EnterpriseOne applications. So where do you start? In this post, I’ll lay down some basic definitions you can work from to start developing your high availability strategy. For more help, just email me at email@example.com or see me at Collaborate ’07. I’ll be presenting on this topic at Session 20710 and representing Global Systems Integration (GSI) at Booth 718.
Let’s Get Started!
What is “high availability” with regards to EnterpriseOne? Well, everyone’s definition of “high” is different, so we should really refer to it as, “HIGHER Availability.” But for the purposes of this blog, the upcoming presentation, and familiarity, we will go with “high availability.” That being said, what is “High Availability for JD Edwards EnterpriseOne?”
“High Availability” for JD Edwards EnterpriseOne is usually a single site implementation that has multiple redundant servers at each layer configured in either an ACTIVE-PASSIVE or ACTIVE-ACTIVE mode to minimize unplanned downtime.”
Let’s break this down a little further so we make sure we all understand what this is really saying.
Why do we say, “…usually a single site…?” This is needed because high availability with multiple sites is usually referred to as “Business Continuity,” and Business Continuity (formerly referred to as the less politically correct, “Disaster Recovery”) is a huge subject on its own. So, to keep the “high availability” concept palatable, we will only consider single sites in our discussion.
The next part of the definition, “…multiple redundant servers…,” really means multiple servers that are configured with the same level of code AND that perform the same function. Although this sounds like a grammatical train wreck, we really wanted as few words as possible to show that there can be multiple (more than two) servers with the same function.
The next phrase, “…at each layer…,” is a GSI concept that is supposed to pictorially depict a segmentation of the different EnterpriseOne server roles. We chose “layers” because we felt this concept was very similar to the seven layers of the OSI Model. (Do you remember the seven layers? Hint: Please Do Not Throw Sausage Pizza Away = Physical, Data Link, Network, Transport, Session, Presentation, Application.) To be completely honest, it is really something we figured all “geeks at heart” and the Configurable Network Computing community could relate to easily.
Note: There will be more on this subject in my presentation, but for now, just note the new way of thinking about EnterpriseOne Servers…layers.
Lastly, what is meant by “ACTIVE-PASSIVE” and “ACTIVE-ACTIVE?”
In ACTIVE-PASSIVE mode, only ONE server at each layer is ONLNE. All other redundant servers at that layer are in STANDBY mode. Upon a failure at any layer, the STANDBY servers switch to ONLINE and the end users are re-routed either MANUALLY or AUTOMATICALLY to the ONLINE servers.
In ACTIVE-ACTIVE mode, ALL servers at ALL layers are ONLINE at ALL times. This means that end users can be spread across multiple servers equally (load balanced) or “weighted” based on a server’s available resources (some servers may be have more CPU/RAM than other servers). Upon failure at any layer, only the users on the failed server will be re-routed to an ONLINE server while the other users will remain unaffected.
Hopefully, you now understand what we mean by…
“High Availability” for JD Edwards EnterpriseOne is usually a single site implementation that has multiple redundant servers at each layer configured in either an ACTIVE-PASSIVE or ACTIVE-ACTIVE mode to minimize unplanned downtime.
We will go into much greater detail, discuss the multiple options that are available to you, and show you a couple of live client configurations during our presentation at Collaborate 07. If you can’t attend, please check back here at the end of April and I will post the slides from the presentation.
If you have any questions about the information provided here or about Load Balancing, Virtualization, Upgrades, CNC, or Business Continuity, please feel free to use the comment box below or contact me at firstname.lastname@example.org. In addition, I will be representing GSI at Booth 718 (ACG-GSI) during Collaborate ‘07, so please stop by for a visit.
Kevin Herrig is the CEO of Global Systems Integration (GSI), a technology consulting firm that specializes in Configurable Network Computing (CNC) and related technologies to JD Edwards OneWorld and EnterpriseOne customers. He welcomes your questions and thoughts at email@example.com.